Normalized Difference Vegetation Index (NDVI) derived from Moderate Resolution Imaging Spectroradiometer (MODIS) time-series data has been widely used in the fields of crop and rice classification. The cloudy and rainy weather characteristics of the monsoon season greatly reduce the likelihood of obtaining high-quality optical remote sensing images. In addition, the diverse crop-planting system in Vietnam also hinders the comparison of NDVI among different crop stages. To address these problems, we apply a Dynamic Time Warping (DTW) distance-based similarity measure approach and use the entire yearly NDVI time series to reduce the inaccuracy of classification using a single image. We first de-noise the NDVI time series using S-G filtering based on the TIMESAT software. Then, a standard NDVI time-series base for rice growth is established based on field survey data and Google Earth sample data. NDVI time-series data for each pixel are constructed and the DTW distance with the standard rice growth NDVI time series is calculated. Then, we apply thresholds to extract rice growth areas. A qualitative assessment using statistical data and a spatial assessment using sampled data from the rice-cropping map reveal a high mapping accuracy at the national scale between the statistical data, with the corresponding R² being as high as 0.809; however, the mapped rice accuracy decreased at the provincial scale due to the reduced number of rice planting areas per province. An analysis of the results indicates that the 500-m resolution MODIS data are limited in terms of mapping scattered rice parcels. The results demonstrate that the DTW-based similarity measure of the NDVI time series can be effectively used to map large-area rice cropping systems with diverse cultivation processes.

Human activities can affect soil nutrients, thereby influencing river water quality. The spatial pattern of precipitation also impacts distributions of water quality. In this paper, we employed a method that combines point survey, soil, and water quality data to analyze the spatial relationships between precipitation, soil nutrient and water quality in the basin on the basis of field surveys and laboratory analysis. The ordinary kriging method was applied to interpolate the precipitation and soil data, and the spatial pattern was analyzed. The water samples on the main stream and soil samples in the field were collected during both the dry and rainy seasons to analyze the water quality and soil nutrients. The results indicate: (1) The water quality in the dry season is better than that in the rainy season, the water quality in the upper reaches is better than that in the lower reaches, and agricultural activity is the direct source of water pollution. (2) The precipitation in the rainy and dry seasons is differente and the dilution effect of precipitation on pollutant concentrations and transport of water flow affect the spatial distribution of water quality. (3) There is a significant difference in the spatial pattern of soil nutrients between the dry and rainy seasons, and the soil nutrient content and the surface runoff directly affect the water quality. Soil nutrients are affected by human activities, and they potentially act as nonpoint source (NPS) pollution in this river basin. To improve the water quality, suitable agriculture measures need to be implemented.

The Yellow River Delta (YRD), located in Yellow River estuary, is characterized by rich ecological system types, and provides habitats or migration stations for wild birds, all of which makes the delta an ecological barrier or ecotone for inland areas. Nevertheless, the abundant natural resources of YRD have brought huge challenges to the area, and frequent human activities and natural disasters have damaged the ecological systems seriously, and certain ecological functions have been threatened. Therefore, it is necessary to determine the status of the ecological environment based on scientific methods, which can provide scientifically robust data for the managers or stakeholders to adopt timely ecological protection measures. The aim of this study was to obtain the spatial distribution of the ecological vulnerability (EV) in YRD based on 21 indicators selected from underwater status, soil condition, land use, landform, vegetation cover, meteorological conditions, ocean influence, and social economy. In addition, the fuzzy analytic hierarchy process (FAHP) method was used to obtain the weights of the selected indicators, and a fuzzy logic model was constructed to obtain the result. The result showed that the spatial distribution of the EV grades was regular, while the fuzzy membership of EV decreased gradually from the coastline to inland area, especially around the river crossing, where it had the lowest EV. Along the coastline, the dikes had an obviously protective effect for the inner area, while the EV was higher in the area where no dikes were built. This result also showed that the soil condition and groundwater status were highly related to the EV spatially, with the correlation coefficients −0.55 and −0.74 respectively, and human activities had exerted considerable pressure on the ecological environment.

Parameters for the tasselled cap transformation (TCT) have been derived for many sensors since 1976. There have been concerns about the comparability of TCT brightness (TCTB), greenness (TCTG), and wetness (TCTW) from different sensors because the number and bandwidth of spectral bands of the different sensors are not exactly the same and the derivation methods vary. In this research, comparisons between the TCT components derived from different combination images with a different number of spectral bands are considered. First, a new TCT based on a new data set from Landsat 8 Operational Land Imager top of atmosphere (OLI TOA) reflectance was developed. Then, a case study of the Yellow River Delta, China, demonstrated that TCT parameters derived from a Landsat 8 OLI TOA reference image from May 2013 were probably applicable to clear and nearly cloud-free images for spring, summer, and autumn over the Yellow River Delta. Finally, we compared the TCT components derived from selected bands of Landsat 8 OLI TOA reflectance images and those derived from images from other well-known moderate-resolution worldwide remote-sensing data by statistical characteristics, correlation coefficients, the optimum index factor (OIF), and classification with a support vector machine. The result supports our conclusion (a) that two new shortwave bands (Bands 1 and 9) of the Landsat 8 OLI have little effect on derivations of the TCT components and their ability to classify land cover type and (b) that Bands 4–7 in Landsat multispectral scanner, Bands 1–4 in Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), and Bands 1–4 in Systeme Probatoire d’Observation dela Tarre 5 are sufficient for deriving TCTB, TCTG, and TCTW components and mapping land cover.

Machine learning comprises a group of powerful state-of-the-art techniques for land cover classification and cropland identification. In this paper, we proposed and evaluated two models based on random forest (RF) and attention-based long short-term memory (A-LSTM) networks that can learn directly from the raw surface reflectance of remote sensing (RS) images for large-scale winter wheat identification in Huanghuaihai Region (North-Central China). We used a time series of Moderate Resolution Imaging Spectroradiometer (MODIS) images over one growing season and the corresponding winter wheat distribution map for the experiments. Each training sample was derived from the raw surface reflectance of MODIS time-series images. Both models achieved state-of-the-art performance in identifying winter wheat, and the F1 scores of RF and A-LSTM were 0.72 and 0.71, respectively. We also analyzed the impact of the pixel-mixing effect. Training with pure-mixed-pixel samples (the training set consists of pure and mixed cells and thus retains the original distribution of data) was more precise than training with only pure-pixel samples (the entire pixel area belongs to one class). We also analyzed the variable importance along the temporal series, and the data acquired in March or April contributed more than the data acquired at other times. Both models could predict winter wheat coverage in past years or in other regions with similar winter wheat growing seasons. The experiments in this paper showed the effectiveness and significance of our methods.